Deep Learning (Zettelkasten)
v1.0.0全能深度阅读工具（All-in-One Deep Reading）。适用于书/长文/研报/论文的深度消化。Use when 用户要深度消化一本书/长文/研报/论文并构建知识网络、产出结构笔记与原子笔记。融合 Mortimer Adler（结构）、Feynman（解释）、Luhmann（网络）、Pragmatist（工具化）、Critics（辩论）；强制 High Fidelity 案例保留与 Actionable 工具提取。关键词：深度阅读、结构笔记、deep learning、卢曼。
⭐ 0· 763·3 current·3 all-time
by@mikonos
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (deep reading, structure/index/atomic/method notes) match the SKILL.md and templates. No unrelated environment variables, binaries, or installs are requested. References to companion skills (structure-note, index-note, file-organize, workflow-audit) are appropriate for automating file/note operations.
Instruction Scope
The runtime instructions explicitly create, move, and modify files and folders in user vault paths (e.g., 05_每日记录/YYYY/MM/DD, 03_索引/). That behavior is expected for a note-management skill but is broad: the agent is authorized by the skill to write and relocate notes. The SKILL.md also mandates calling other skills when present. Users should be aware this will change filesystem state and could overwrite/move notes if run blindly.
Install Mechanism
No install spec and no code files — instruction-only. Lowest-risk install footprint (nothing downloaded or executed beyond the agent running its normal actions).
Credentials
No environment variables, credentials, or config paths are requested. All file path assumptions are declared in README and SKILL.md; requested access is limited to note directories and companion skills, which is proportionate for a vault-management skill.
Persistence & Privilege
always:false (no forced inclusion). The skill assumes the agent may autonomously invoke it (default platform behavior). Because the instructions perform file creation/movement and can call other skills (file-organize, workflow-audit), autonomous invocation increases the blast radius — this is expected for automation but users should confirm they want automatic file-editing.
Assessment
This skill is internally consistent and doesn't request secrets or install code, but it will create, move, and edit files in your vault (e.g., 05_每日记录 and 03_索引). Before installing or letting the agent run it autonomously: 1) Back up your notes or test in a copy of your vault. 2) Confirm the path names in the skill match your folder structure (or edit the SKILL.md/templates). 3) If the skill calls companion skills (structure-note, index-note, file-organize, workflow-audit), verify you trust those skills — they may also modify files. 4) Run the workflow manually once to inspect outputs before enabling fully autonomous runs. If you want extra caution, disable autonomous invocation for this skill or require explicit user confirmation before performing file operations.Like a lobster shell, security has layers — review code before you run it.
latestvk974tj179v01nyy3bmp1sr5yh5811djg
License
MIT-0
Free to use, modify, and redistribute. No attribution required.